What big ideas have you identified in this chapter?
CiCi's Sunday January 23rd. 2-4
Subscribe to:
Post Comments (Atom)
Email: jean.linner@cobbk12.org
Extra help is available before and after school. See you at CiCi's!
Lassiter High School
2601 Shallowford Road
Marietta, GA 30066
770-494-7863
Copyright 2008 Cobb County School District. All rights reserved.
Submit your answers to Columbus State's POW
About Me
- Mrs.L
- AP Statistics info for Lassiter students and our friends. Email and webmaster:jean.linner@cobbk12.org
Posts
19 comments:
What are you using for your Freq: in Stat Plot? It should be 1 because your raw data are in L1.
Be sure to note whether the cursor is in ALPHA or regular mode before you try to change the Freq. To turn off the ALPHA mode, just hit the ALPHA key.
I am so happy for you!
The radian/degree setting will not affect your results. I am glad that you are using radians in your other math class. Radians are sooooooooooo useful, even though students are much more familiar with degrees and shy away from radians.
Way to go!
Find P( .5 < p-hat) by standardizing both .5 and p-hat.
You'll have P( (.5-.52)/sqrt(.52*.48/500) < z). That's where the - .8951 comes from.
Now, in a standard normal distribution, what is the area to the right of -.8951?
Essentially, you are using a normal distribution to make calculations that would be binomial a whole lot easier.
Thank you for asking!
First, the population is the "large population" the author cited in the problem. The samples are the two groups of 2000 men.
So far in this section we have only worked problems where we are investigating probabilities. Therefore, can we convert 75 heart attacks out of 2000 men to a percentage? That is the p-hat value that you are comparing to .04. Follow the procedures we saw in class today: standardizing, finding the area under the curve, writing up the results.
Preview of the next topic: If we wanted to compare X=75 to our mu of 80 we'd have to change our procedure a little bit. The first problem is determining the standard deviation of X. We'll work on this a little later.
Good job of problem-solving! I believe that interpreting will get easier with all of your hard work.
When asked for the distribution you should identify the center, spread, and shape. What is the mean of the three weighings? What is the standard deviation of the averages you would get? What type of distribution is this? Why are you sure that the sampling distribtion is normal, even though the sample size is so small?
Twelve is the sample size. What is the average egg weight for your sample of 12 eggs?
BTW-GG is a repeat tonight.
You are my favorite students!!!!
Let's see, in order. . .
for problem 9.42, what is the sample size? Other than that, your methods look good. This is a problem where you are working with p-hat, the proportion of the sample/population representing high school dropouts.
For problem 9.44, you are looking at the average weight of ONE egg, using a sample of size 12. The std dev of the sample average weight is sigma x /sqrt sample size. That is where the 5/sqrt 12 came from. This is a problem where you are working with the mean of X.
You ladies make me proud.
In every problem we use the same basic form, but the values we plug in look a little different.
The form is z = (STATISTIC - PARAMETER)/STD DEV OF STATISTIC.
For problems with means, use
z = (x-bar minus pop mean)/(std dev of x-bar).
For problems with proportions, use
z = (p-hat minus p)/(std dev of p-hat).
What is a sampling distribution?
How does the distribution of x differ from the distribution of x-bar?
What is bias?
What is variability?
What effect does increasing or decreasing the sample size have on the variability?
How do you know if you can use the normal approximation?
How do you know if you can use the formula for the std dev of p-hat?
How do you interpret results?
How can you check your answers to probability-related questions quickly on the calculator? (but this is not acceptable work for a free response question!)
What is a "parameter of interest"?
What is a "statistic of interest"?
kcorky--the correct answer is B. I hope that everyone can figure out how to calculate this answer. The questions on the online review are nice and challenging.
jenny--re: your question--if you are given x and n INSTEAD of a p-hat value, then x/n is how you will calculate the p-hat. You won't be given BOTH p-hat and a target number of successes (x).
Re: when can you use the normal approximation--There's a little more to it. What do the two rules of thumb allow you to do? Also, what are the rules about x values that aren't normally distributed? When can x-bar still be approximately normal if the x values themselves are wild and crazy?
Caitlin- #4 -the problem doesn't involve z OR x-bar OR mu. Just use the formula for the standard deviation of p-hat.
#10 says that there is a point somewhere above the mean where 99% of the x-bars will be below it. Find that point, then find out how far above the mean it is.
Lassiterbaseball11 - In that case you'd have to calculate the probabilities directly using binomial formulas (or BinomCDF on the calculator).
Guys--
When you solved
.01 = sqrt(.2*.8/n) for n you got n=16. That is NOT the variance. What is n?
Did any of you see that your most excellent colleague told us the answer to # 4 was wrong and that I confirmed it?
Look higher in the postings.
Problem 7:
There's only one way that the average could be more thn $65.
.5 * .5 = .25
Problem 10:
I already provided an interpretation of this earlier in the blog.
Find the value L such that there is a probability of .99 that the average duration of the 30 patients lies less than L yers above the overall mean of 8 yeaars.
L is the difference between the value of x-bar, a point somewhere higher than mu, and mu.
Start by finding the z value that leaves only 1% above the value. (InvNorm(.99))
Sorry. I CAN spell. I just can't type.
Number 8--
This goes back to my original questions. If the underlying values of x are normally distributed, then the values of x-bar that you calculate from the samples are all EXACTLY normally distributed.
If the underlying data were only mound-shaped, then you would need a moderate size sample to make the x-bar values approximately normally distributed.
If the underlying data were distinctly non-normal, then you would need a LARGE sample size for the x-bar values to be approximately normally distributed.
All of these concepts are cleverly hidden in our text, in the boxes that I read to you. Please confirm this in your resources.
LassiterLaxChick--
Nice job explaining #9.
P(7 < x-bar < 9)
What is the area under the normal curve between 7 and 9 when the mean is 8 and the standard deviation of x-bar is 4/sqrt 30?
Have you guys tried verifying answers like this using the following entries on your calculator. . .
NormalCDF(7,9,8,4/SQRT30)
This will verify answers, but it does not show sufficient understanding for free response problems.
this is an abundance of comments
Post a Comment